1. Field of the Invention
The present invention relates to an image-pickup apparatus and a method of controlling the same.
2. Description of the Related Art
Recently, solid-state image pickup apparatuses are quite popular, each of which uses a solid-state image pickup element, such as a CCD image sensor or a CMOS image sensor. The solid-state image pickup element is applied to a digital still camera, a digital video camera, or the like, and is widely used.
It is known that a phenomenon so called a “fluorescent light flicker” occurs when an object is taken as a moving image under an illumination of a fluorescent lamp by the image-pickup apparatus, such as the digital still camera and the digital video camera. For example, when an image is taken by an image-pickup element having a rolling shutter characteristic under the illumination of the fluorescent lamp directly lit by a commercial alternating current (AC) power, a luminance flicker, so called a “fluorescent light flicker”, may occur in a time direction in each image. This is a phenomenon in which periodical bright and dark patterns occur in an image signal taken in the image-pickup apparatus due to a difference between a frequency twice as much as that of the commercial alternating current power and a vertical synchronizing frequency of the image-pickup apparatus because the fluorescent lamp, as a light source, changes with the frequency twice as much as that of the commercial alternating current power.
Therefore, in order to solve the problem, many method of removing the flicker are proposed, and Japanese Patent Laid-Open No. 2009-17213 obtains a vertical intensity distribution by integrating an output of the CMOS image sensor in a horizontal direction, and calculates a flicker component in a vertical direction in a present frame on the basis of the vertical intensity distribution in a plurality of frames. An original image (taken image before the correction) is corrected by calculating a flicker correction value from the calculated flicker component and by multiplying a video signal of the present frame by the correction coefficient.
In addition, a hue change (color flicker) of an image may occur when only a luminance signal is corrected because, in general, a constituent ratio among R (red), G (green), and B (blue) in the illuminating light of the fluorescent lamp slightly changes according to the luminance of a illumination. Accordingly, Japanese Patent Laid-Open No. 2008-109370 proposes a method of separately correcting R, G, and B flickers instead of correcting only the luminance signal.
Japanese Patent Laid-Open No. 2008-92334 indicates a ratio among R, G, and B, which are color elements of the fluorescent lamp, and points out that a color flicker occurs when a high-speed shutter is used.
A part of an image-pickup area of an image-pickup element may be used as a function other than taking images. Japanese Patent Laid-Open No. 2000-156823 discloses an image-pickup apparatus configured to have a focus detection function by using an image-pickup element including both of a normal pixel (image-pickup pixel) and a focus detection pixel that performs a pupil divide by providing a light-shielding layer between a microlens and a photodiode.
Japanese Patent Laid-Open No. 2000-156823 gathers signals only from focus detection pixels, which are pupil-divided into right and left or up and down, forms two line images which are pupil-divided, and calculates a defocus amount based on a shift amount between the images.
The shift amount between projection positions of the images which are pupil-divided with a certain defocus amount can be determined by the base length between the two divided pupils. The shift amount becomes larger as the base length becomes longer, and the accuracy of the focus detection improves.
In an image-pickup element including both of the focus detection pixel and the image-pickup pixel, the above-mentioned fluorescent light flicker affects the focus detection pixel by its characteristic, and the bright and dark patterns occur. Since the fluorescent light flicker causes an uneven luminance in the vertical direction, the bright and dark pattern occurs in two line images pupil-divided by gathering the signals from the vertically pupil-divided focus detection pixels, and therefore the shift amount between the images cannot be calculated accurately. Accordingly, the defocus amount cannot be calculated accurately.
Since the number of focus detection pixels is less than that of image-pickup pixels or the focus detection pixels are arranged around the center, the flicker cannot be detected simply with the focus detection pixels in the conventional flicker detection method.
As described in Japanese Patent Laid-Open No. 2010-78856, the focus detection signal of the focus detection pixel is not used for an image generation, and therefore a filter of a transparent (white) film may be arranged instead of R, G, and B color filters.